Predicting the Personal-Best Times of Speed Skaters Using Case-Based Reasoning

Speed skating is a form of ice skating in which the skaters race each other over a variety of standardised distances. Races take place on specialised ice-rinks and the type of track and ice conditions can have a significant impact on race-times. As race distances increase, pacing also plays an important role. In this paper we seek to extend recent work on the application of case-based reasoning to marathon-time prediction by predicting race-times for speed skaters. In particular, we propose and evaluate a number of case-based reasoning variants based on different case and feature representations to generate track-specific race predictions. We show it is possible to improve upon state-of-the-art prediction accuracy by harnessing richer case representations using shorter races and track-adjusted finish and lap-times.